CAXADA PORCUPINE 75 



creased to nearly 37 minutes. In this test he was lost chiefly 

 in the second half. On the third trip, or the first one the fol- 

 lowing day, his time was down to two and one -half minutes 

 with four en-ors ; but he did not reach his former ability until 

 the 83d trip, or after 30 trials. This, strange to say, is within 

 one of being the exact number of trials necessary for him to 

 reach the record of a single error in the beginning. It looks 

 like a case of relearning the maze though the curve of releaming 

 is much more regular as will be seen in the figure. Before the 

 84th trial (C plate IV) the maze was again rotated through 

 90 degrees clock-wise. The animal was again completely lost 

 in the same maze. This time the porcupine missed 7 and re- 

 quired 51 minutes and 43 seconds to reach the center. On 

 the second trip that day, however, he required only five minutes 

 and attained his former ability in eight trials as compared with 

 30 trials when the maze was first turned. Before the 92nd 

 test the maze was again turned 90 degrees or 270 degrees from 

 the initial position and the animal was through in seven minutes 

 as compared with nearly 52 minutes when the rotation was 

 made before (D plate IV). He reached his one-error standard 

 in ten trials. The maze was then turned to its original position 

 for the 102nd trial. All went well without an error to 6. He 

 dropped into 6 but withdrew at once. On leaving that point 

 there occurred what is considered the psychological moment of 

 an important choice. The porcupine on this occasion, hesitated 

 with abortive starts between the forward and backward trail 

 six times. Had he gone right it is reasonable to infer that he 

 would have made his single error, and it would likewise have 

 been a reasonable inference that he knew the maze in the posi- 

 tion from the old association. His behavior did reflect earlier 

 experience very strongly, but this time he chose the back trail 

 and wasted nearly 14 minutes in the first half of the maze which 

 he had just previously traversed without difficulty. His time 

 was 14 minutes 55 seconds and his errors 30, but he reached 

 his former record in six trials (E plate IV). Two weeks later 

 with the same maze in the barn and rotated 180 degrees from 

 its last position he found his way in nine minutes and again 

 in three minutes. Similar figures represent the iioth and iiith 

 trips the next day (F plate IV). h 



Watson (44) found that rotating his maze through 180 degrees 



